Ementa Big Data for Finance Project

*For portuguese-br version see the pdf file here.

📌 General Information

  • Course: Data Science for Business
  • Subject: Big Data for Finance Project
  • Workload: 90 hours
  • Programming Languages: R or Python

🎯 Course Description

The Big Data for Finance Project course aims to equip students with the necessary skills to develop applied financial projects using big data analytics techniques.
Students will be encouraged to work with R or Python to build analytical models and visualizations, implementing solutions applicable to real-world contexts.
By the end of the course, each student will have created a GitHub Pages site to publish their projects, forming an individual portfolio for the job market.

🎯 Objectives

  • Apply Big Data concepts to real financial projects.
  • Develop hands-on skills in R or Python for financial data analysis.
  • Create analytical and visualization solutions for financial problems.
  • Build an online portfolio with posts on GitHub Pages.

🔥 Skills Developed

  • Handling and analyzing large-scale financial data.
  • Constructing predictive and diagnostic models.
  • Data visualization for decision-making.
  • Presenting results professionally.

📚 Course Syllabus

️Our first meeting

  • Didatic contract
  • Content of course presentation
  • Justificative for financial ML/DS/Econometrics portfolio

1️⃣ Creating GitHub Pages with Quarto

  • Setting up Quarto
  • Publishing pages on GitHub Pages

2️⃣ Generating Synthetic Data

  • Introduction to synthetic data
  • Methods and tools for simulated data creation

3️⃣ Feature Engineering & Key Metrics for Financial Time Series

  • Dimensionality reduction techniques
  • Building differentiated variables
  • Techniques for stationarity analysis

4️⃣ Forecasting Price Series in Basic Economic Theory

  • Forecasting fundamentals
  • Practical applications in price prediction

5️⃣ Markowitz Theory & Multi-Objective Modeling

  • Introduction to Markowitz model
  • Implementing a bi-objective model for risk-return balance

6️⃣ Considering Investor Profiles in Decision Making

  • Adjustments based on investor risk profiles
  • Strategies for personalized financial decisions

🏆 Teaching Methodology

This course will be practical, including: ✔ Lectures and case studies.
✔ Hands-on lab sessions using R/Python.
✔ Individual and group projects.
✔ Guidance on publishing results in GitHub Pages.

📝 Evaluation Criteria

Activity Weight
Participation & practical exercises 30%
Intermediate project deliveries 30%
Final project published on GitHub Pages 40%

Course grapho

mindmap
  root((Big Data for Finance Project))
    subtopic Introduction ["Creating GitHub Pages with Quarto"]
      subsubtopic ["Initial Quarto setup"]
      subsubtopic ["Publishing pages on GitHub Pages"]
    subtopic Synthetic Data ["Generating Synthetic Data"]
      subsubtopic ["Introduction to synthetic data"]
      subsubtopic ["Methods and tools for generating simulated data"]
    subtopic Feature Engineering ["Feature Generation for Financial Time Series"]
      subsubtopic ["Building differentiated variables"]
      subsubtopic ["Techniques for analyzing stationarity"]
    subtopic Forecasting ["Forecasting Price Series in Economic Theory"]
      subsubtopic ["Fundamentals of forecasting"]
      subsubtopic ["Practical applications in price forecasting"]
    subtopic Markowitz ["Basic Markowitz Theory and Multi-objective Modeling"]
      subsubtopic ["Introduction to Markowitz Model"]
      subsubtopic ["Bi-objective modeling for return and risk"]
    subtopic Decision Making ["Considering Investor Profile in Decision-Making"]
      subsubtopic ["Adjustments based on investor profile"]
      subsubtopic ["Strategies for personalized decisions"]

Code 
# Load required libraries
library(igraph)
library(networkD3)
library(dplyr)

# Create dataset following the structure of the Mermaid mindmap
data <- data_frame(
  from = c(
    "Big Data for Finance Project", "Big Data for Finance Project", "Big Data for Finance Project",
    "Big Data for Finance Project", "Big Data for Finance Project", "Big Data for Finance Project"
  ),
  to = c(
    "Creating GitHub Pages with Quarto", "Generating Synthetic Data", 
    "Feature Generation for Financial Time Series", "Forecasting Price Series in Economic Theory",
    "Basic Markowitz Theory and Multi-objective Modeling", "Considering Investor Profile in Decision-Making"
  )
)

# Expand the dataset to include subtopics
subtopics <- data_frame(
  from = c(
    "Creating GitHub Pages with Quarto", "Creating GitHub Pages with Quarto",
    "Generating Synthetic Data", "Generating Synthetic Data",
    "Feature Generation for Financial Time Series", "Feature Generation for Financial Time Series",
    "Forecasting Price Series in Economic Theory", "Forecasting Price Series in Economic Theory",
    "Basic Markowitz Theory and Multi-objective Modeling", "Basic Markowitz Theory and Multi-objective Modeling",
    "Considering Investor Profile in Decision-Making", "Considering Investor Profile in Decision-Making"
  ),
  to = c(
    "Initial Quarto setup", "Publishing pages on GitHub Pages",
    "Introduction to synthetic data", "Methods and tools for generating simulated data",
    "Building differentiated variables", "Techniques for analyzing stationarity",
    "Fundamentals of forecasting", "Practical applications in price forecasting",
    "Introduction to Markowitz Model", "Bi-objective modeling for return and risk",
    "Adjustments based on investor profile", "Strategies for personalized decisions"
  )
)

# Combine both datasets
full_data <- bind_rows(data, subtopics)

# Plot the interactive network graph
p <- simpleNetwork(full_data, height = "600px", width = "100%",        
                   Source = 1,                # Source column
                   Target = 2,                # Target column
                   linkDistance = 80,         # Distance between nodes
                   charge = -500,             # Repulsion/attraction strength
                   fontSize = 14,             # Font size of nodes
                   fontFamily = "serif",      # Font type
                   linkColour = "#666",       # Edge color
                   nodeColour = "#69b3a2",    # Node color
                   opacity = 0.9,             # Node transparency
                   zoom = TRUE                # Enable zooming
)

p